Autism is a devastating neuropsychiatric condition with unknown pathophysiology. Autism spectrum disorders (ASD) have an estimated incidence of 1/200 and thus are more common than many other childhood disorders. Although ASD have a multifactorial etiology, it has a large genetic component. It is also becoming clear that comprehensive efforts involving large sample sizes and methods to reduce heterogeneity are necessary to achieve maximal power to identify disease critical regions narrow enough to permit positional cloning of autism susceptibility genes. The investigators in this application aim to continue their collaborative effort that has produced and enhanced a highly successful open data and biomaterials resource for the research community, the Autism Genetic Resource Exchange (AGRE). This collaborative network application involving six research sites and the AGRE DCC, will systematically and comprehensively investigate the genetics of ASD to identify rare mutations, chromosomal abnormalities, and common variation contributing to ASD susceptibility. Specifically, they will enrich existing resources by adding 400 simplex families, including 200 families of African American descent, in addition to phenotype enrichment. The large overall sample size permits stratification of families based on analysis of heritable quantitative and qualitative endophenotypes. It further allows independent confirmation of loci as demonstrated for chromosome 17 and provides adequate sample for whole genome association studies to find loci with adequate power. The investigators will perform follow up linkage studies to confirm several new loci identified based on autism-related endophenotypes or co- variants, such as language delay, sex, and head circumference. In parallel, comparative genomic hybridization (CGH) using 500k SNP arrays will be performed, yielding the highest resolution molecular karyotypes and providing a resource on genome wide copy number variation (CNV) in ASD. CNV identified will be followed in family members and controls using QPCR and FISH. The use of SNP genotyping for CNV detection in ASD probands further provides for significant economies, since additional genotyping need only be conducted in the parents for efficient whole genome association studies. Regions or genes identified by linkage will be followed up by efficient, staged dense SNP genotyping. CNV assessment and WGA will also yield candidates that will be integrated with the linkage results for focused confirmatory studies, including re-sequencing to identify and confirm potentially causal genetic variants. Genetic risk factors identified in the mostly white European sample will be tested for association in the African American sample to determine whether these cohorts share the same genetic risk factors. All phenotypic and genotype data will be made accessible via the Internet on a rolling basis, including minority families, further enhancing the value of this resource to the community. [unreadable] [unreadable] [unreadable]